5. CONCLUSIONS
3.4. Thermodynamic modelling and P-T paths
samples collected 106 to 1840 m below the fault. This equates to thermal gradients from 50 to 1100 ˚C / km. Calculated thermal gradients increase with decreasing exposed structural thickness from east to west (Figure 3.2), suggesting that structural attenuation is at least partly responsible for the 1100 ˚C / km value. Thermal gradients of 50 to 240 ˚C / km (calculated from eastern and central transects, respectively) are similar to those reported in the schist of Sierra de Salinas (50 to 70 ˚C; Kidder and Ducea, 2006) and in the Pelona Schist (240 ˚C / km; Graham and Powell, 1984).
Figure 3.10. Calculated “prograde” P-T pseudosections for samples (a) 06SE23 and (b) 08SE479B. THERMOCALC results overlain for comparison. Numbered fields listed in Appendix 1. Bulk compositions listed in A1.6.
Figure 3.11. Pseudosection shown in Figure 3.10a with contours of (a) volume percent H2O in solids, (b) volume fraction garnet and (c-f) garnet composition expressed as mol fractions overlain. Arrow represents inferred prograde P-T trajectory.
strongly pressure-dependent, and at high temperatures Xgrs is as pressure-dependent as Xsps. We calculate conditions of garnet nucleation from the intersection of end member isopleths that correspond to garnet core compositions (e.g. Vance and Mahar, 1998). The intersection of the garnet core composition isopleths (Xalm = 0.35, Xsps= 0.37, Xgrs= 0.26, and Xprp= 0.02) indicates the temperature and pressure conditions at the onset of growth. Although garnet contains no inclusions useful for thermobarometry, which prohibits an independent check of the conditions of garnet core formation, the tight intersection of the compositional contours at the point 4.3 kbar, 475 ˚C strongly suggests that the pseudosection accurately reproduces the physical conditions of garnet nucleation.
Mafic sample 08SE479b contains the peak assemblage garnet + plagioclase + hornblende + quartz and is representative of mafic schist outcrops > 100 m from the Rand fault. The pseudosection of Figure 3.10b is overlain by garnet isopleths in Figure 3.12.
Garnet core composition isopleths (Xalm = 0.36, Xgrs=0.35, Xsps=0.28, and Xprp=0.02) intersect at approximately 6.5 kbar, 575 ˚C. Mineral isopleth-based constraints of garnet nucleation in both metasandstone and mafic schist lie well inside the stability field of garnet as predicted by equilibrium growth (Figures 3.10, 3.11, and 3.12). This is most likely due to the effect of missing Mn end-members in the solution models of competing phases. The presence of Mn-bearing chlorite, biotite, and ilmenite during prograde metamorphism would hinder garnet nucleation and growth at low metamorphic grades, restricting its stability field to higher pressures and temperatures. Additional explanations for this observation include:
1) late metastable nucleation of garnet (i.e., overstepping of the garnet isograd), 2) limited local equilibrium due to slow reaction kinetics at low temperatures, and 3) diffusional modification of the core.
Figure 3.12. Pseudosection shown in Figure 3.10b with contours of (a) volume percent H2O in solids, (b) volume fraction garnet and (c-f) garnet composition expressed as mol fractions overlain. Arrow represents inferred prograde P-T trajectory.
Prograde paths for garnets in samples 06SE23 and 08SE479b are approximated by interpolating the conditions of garnet core formation with peak metamorphic conditions.
The preferred prograde paths outlined in Figures 3.11 and 3.12 incorporate these constraints and take into consideration 1) the lack of prograde clinozoisite in sample 06SE23 and the presence of prograde epidote in sample 08SE479b; 2) the preservation of primary muscovite in sample 06SE23; and 3) the lack of evidence for partial melting in these samples, although such evidence is present at higher structural levels in the schist. Therefore, the likely prograde path for sample 06SE23 involves burial from approximately 4.3 to 10 kbar and contemporaneous heating from 475 to 665 ˚C. Along this path, the only net transfer reaction encountered in sample 06SE23 during garnet growth is the chlorite-out reaction. The lack of primary chlorite as inclusions within garnet probably resulted from garnet scavenging Fe, Mg, Al, and Si from the mineral, which was likely present during garnet core growth. The prograde trajectory of sample 08SE479b overlaps that of 06SE23, with conditions increasing from 7 kbar, 575 ˚C to 8.4 kbar, 636 ˚C. It is important to note that although the preferred prograde P-T paths shown in Figures 3.11 and 3.12 explain many core-to-rim garnet zonation trends in samples 06SE23 and 08SE479b (e.g. monotonically increasing Xprp, monotonically decreasing Xsps, approximately constant Xgrs, and increasing Xalm prior to reaching a compositional plateau in the outer 400 µm), modelled garnet isopleths do not match the composition of garnet beyond the cores at any point in P-T space. In other words, the garnet core compositions are the only regions that can be successfully modelled. This is because the assumption of equilibrium garnet growth is no longer valid outside the cores.
Isopleth intersection thermobarometry breaks down following garnet core growth because the bulk rock composition is changed as garnet sequesters material, particularly Mn, forever
Figure 3.13. Comparison of modelled garnet zonation patterns (a-c) and observed garnet compositional profiles (d). Modelled profiles for the prograde P-T trajectory shown in Figure 3.11 (sample 06SE23) for (a) homogeneous equilibrium crystallization, (b) 10%
fractional crystallization, and (c) 25% fractional crystallization.
removing it from the system unless the temperature becomes sufficiently high that diffusion and/or resorption reintroduces it to the system (Spear, 1993; Vance and Mahar, 1998).
Figure 3.13 compares modelled and observed garnet zonation patterns and illustrates the effect of fractional crystallization for the preferred prograde path of sample 06SE23 (see Appendix 1 for a description of the calculation routine). This effect is to deplete the bulk rock composition, leading to a decrease in nutrient availability at the garnet surface and a corresponding decrease in growth rate. While the observed garnet zonation is fit quite well by the equilibrium growth model, some proportion of fractional crystallization may explain sharp compositional gradients in San Emigdio Schist garnets, particularly in Mn. While we do not consider the effect of volume diffusion in garnet, some amount of diffusional relaxation of these steep compositional variations has likely occurred. High-Ca annuli in garnets from metasandstone schist (Figures 3.5 and 3.13) cannot be explained by fractional crystallization alone and likely reflect chemical disequilibrium during prograde metamorphism (Chernoff and Carlson, 1999; Konrad-Schmolke et al., 2005; Chapman et al., 2006).
Zonation preserved in garnets from samples 06SE23 and 08SE479b indicates that growth did not occur under decreasing pressure and/or temperature conditions (i.e., following peak conditions). Contours of garnet volume % are generally nearly isothermal with slightly negative slopes, indicating that it is impossible to grow garnet during decompression in these bulk compositions. Furthermore, garnet rims are characterized by decreasing Xalm and Xsps and increasing Xprp, which is only possible under conditions of increasing temperature (Figures 3.11 and 3.12).
The prograde path described above involves dehydration reactions, fluid loss (Figures 3.11a and 3.12a), and sequestration of Al, Ca, Fe, Mg, Mn, and Si associated with garnet fractionation. In addition, peak assemblages are well preserved, indicating that the necessary H2O was not available to promote extensive retrogression. However, free water must have been locally present in order to stabilize the D2 assemblage. Therefore, in order to reconstruct the retrograde path of the schist, separate pseudosections were drawn for the bulk compositions of samples 06SE23 and 08SE479B with sufficient free water (i.e., water saturated) to stabilize chlorite and clinozoisite along the retrograde path (Figure 3.14). For simplicity, we assume that the bulk composition effectively equilibrating during retrogression is similar to the original bulk composition from which garnet was subtracted (garnet rims preserve peak P-T composition and show no retrogressive modification). We estimated the modal proportion of garnet, approximated the composition of garnet grains by subdividing them into constituent zones and adding together the contribution of each zone, and finally calculated and subtracted the contribution of garnet to the bulk compositions.
The retrograde paths outlined in Figure 3.15 are based on the stabilization of D2 assemblages. As garnet growth ceased and metasandstone schist (06SE23) entered the retrograde portion of the P-T path, chlorite and clinozoisite were stabilized. According to the retrograde pseudosection for sample 06SE23, this sample must have followed a trajectory with similar P-T slope during decompression than during burial in order to pass through the stability fields of chlorite and clinozoisite. Although epidote entered the stable mineral assemblage in mafic schist (08SE479b) varieties following peak metamorphism, chlorite and actinolite stability fields were not traversed along the retrograde path. The pseudosection for sample 08SE479b indicates that this is possible if the retrograde P-T
Figure 3.14. Calculated “retrograde” P-T pseudosections for samples (a) 06SE23 and (b) 08SE479B with (c and d) contours of volume percent H2O in solids overlain. Numbered fields listed in Appendix 1. Bulk compositions listed in Table A1.6.
Figure 3.15. P-T evolution for samples (a) 06SE23 and (b) 08SE479B. Prograde paths shown as solid lines. (a) “Core” conditions from intersection of garnet isopleths in Figure 3.11, “rim” conditions from Figure 3.9, end-member retrograde paths (dashed lines) constrained by the stability of chlorite and clinozoisite (fields from Figure 3.14a). (b) “Core”
conditions from intersection of garnet isopleths in Figure 3.12, “rim” conditions from Figure 3.9, end-member retrograde paths (dashed lines) constrained by the stability of clinozoisite and the lack of retrograde chlorite (fields from Figure 3.14b).
trajectory passed between chlorite/actinolite-out and clinozoisite/epidote-out reaction boundaries. Figure 3.15 shows, for both metasandstone and mafic schist, two end-member retrograde paths that satisfy the constraints outlined above: 1) one with a shallower P-T slope than the prograde path (i.e., a counterclockwise P-T loop) and 2) one with a steeper slope (i.e., a clockwise P-T loop). While the retrograde path cannot be further constrained, we consider it likely that the geometry of the P-T loop for the schist lies between the end- member possibilities with roughly overlapping prograde and retrograde P-T trajectories (i.e.,
“Franciscan-type” or “out and back” P-T path; Ernst, 1988).
The retrograde paths shown in Figure 3.15 involve the formation of clinozoisite/epidote and/or chlorite, which require the presence of H2O in order to take place. Along these retrograde paths, each sample encounters contours of increasing H2O content, implying that the assemblage becomes fluid-absent unless fluid is available along grain boundaries or is externally derived (Guiraud et al., 2001) (Figure 3.14). Therefore, overprinting of peak assemblages by D2 minerals is evidence for open system behavior involving H2O influx.